Le 
The phases in the history of the apparatus admit of easiest eluci- 
dation if some of the conditions in a limited number of cases be 
chosen for description: and a few of the more striking facts will now 
be picked out, beginning with some of those which have been made out 
in an embryo Raja batis of about 9 mm in length‘). 
Soon after the wandering of the spinal ganglionic foundations 
from the dorsal lips of the cord has taken place, — or even before 
this process is completed — one sees that, in a region extending from 
about the 6 to the 31* trunk somite inclusive, in nearly every trans- 
verse section a few of the most dorsally situated cells of the spinal 
cord have begun to take on ganglionic characters. They grow large 
and the nucleus increases much in size. Owing to this process the 
roof of the cord, as seen in transverse section, looks as though it 
had undergone an upheaval, and presents an appearance quite different 
from that described at corresponding stages in Torpedo”). 
Anteriorly, from segments 6 to 10, these cells are few in number, 
then, from segments 11 to 26, they come in crowds of 3, 4 or more 
to each side of the body in nearly every transverse section, intervals 
of crowding alternating with others in which they are more sparsely 
collected. Finally they diminish from the 26 segment backwards, 
and disappear about the region of the 31% trunk metamere *). 
In a former description of the few such cells occasionally met 
with in Mustelus vulgaris their abnormal situation in the meso- 
derm was commented upon. It would perhaps be more correct to say 
that their occurrence at all in this form is abnormal; for but isolated 
individuals were noticed, and then not in every embryo examined *). 
We must now proceed to disrobe these few cells of Mustelus of 
1) 86 protovertebrae or somites and 5 gill-clefts were present. 
2) The development of the peripheral nervous system in Vertebrates. 
Pt. I. Quart. Journ. Microse. Sci., Vol. 29. 
3) Although, as stated elsewhere (loc. cit.) these big cells are pretty 
evenly distributed over the whole length of the cord in Lepidosteus, 
Seyllium etc., it is remarkable that in Raja (batis, clavata, and 
radiata) they are almost entirely absent in regions of the nervous 
system other than the above specified. This statement holds only for 
specimens under 2 centimetres in length, for they are to be met with 
in the caudal region of examples of Raja batis 5—9 centimetres and 
upwards in length. 
4) Prof. P. Mayer found none in Mustelus, and no wonder, for 
unless the examination be restricted to horizontal sections, it is a tiresome 
task hunting for them here. 
13* 
